MOLECULAR MECHANISMS OF NEURODEGENERATION IN DNA REPAIR DISEASES[unreadable] [unreadable] Cyclopurines as Candidate Neurodegenerative DNA Lesions in Xeroderma Pigmentosum [unreadable] I have proposed that the accumulation of cyclopurine lesions on the transcribed strand of active genes is responsible for the neurodegeneration observed in patients with xeroderma pigmentosum who lack the capacity to carry out NER (Brooks, 2007). [unreadable] To investigate the possibility that RNA polymerase II can bypass cyclopurines in living mammalian cells, we developed a novel method to detect transcriptional mutagenesis in human cells (Marietta and Brooks, 2007). Using this approach, we found that two different classes of mutant RNA transcripts are produced when the polymerase bypasses the cyclopurine lesion in vivo. In the first class, the polymerase incorporates UMP opposite the cyclo-dA lesion, then misincorporates an A opposite the next nucleotide downstream (also an A). We refer to these as 5A mutations. In the second class of mutations, which we call MNDs (multiple nucleotide deletions), the polymerase incorporates U opposite the lesion, and then reinitiates transcription at a location 7, 13, or 22 nucleotides downstream. Notably, we also detected 12 nt MNDs using plasmid DNAs containing a cis-syn thymine dimer (TT-dimer) lesion as well. The TT-dimer lesion had previously been thought to be a complete block to transcription by RNA polymerase II. These results represent the first evidence that DNA lesions of the type that are repaired by the NER pathway can stimulate transcriptional mutagenesis. [unreadable] In terms of mechanism, we observed both types of transcripts in cells lacking the XPD protein or CSB protein, indicating that neither protein is essential for the formation of either type of transcript. However, the ratio of MND/5A mutations was significantly lower in CSB-deficient cells that in the other cell types, indicating that the CSB protein can modulate the probability of MND transcript formation. To explain these unexpected results, we proposed a testable model of how the RNAPII can gene rate transcripts containing MNDs (Marietta and Brooks, 2007).[unreadable] A Minimal system to study DNA lesion effects on RNA polymerases[unreadable] To better understand the molecular mechanism underlying the formation of mutant transcripts, we have recently adopted a minimal system for studying the effects of cyclopurine (as well as other DNA lesions) on transcription by purified RNA polymerases in vitro. Using the minimal system we have confirmed the blocking effect of the cyclopurine lesion on transcription by both mammalian and bacterial RNA polymerases. Experiments using single rNTPs have shown that the polymerase specifically incorporates U opposite the cyclo-dA lesion, consistent with our in vivo observations. We have also extended the minimal system approach to single subunit RNAPs, using T7 RNAP as a model for mitochondrial RNAP. We found that the cyclo-dA lesion is also a block to T7RNAP as well, and using this enzyme, only UMP is incorporated opposite the lesion. [unreadable] [unreadable] Cyclopurines as Inhibitors of mismatch repair [unreadable] We have also explored the ability of cyclo-dA and another oxidized adenosine lesion, 8-oxo-dA, to interfere with 5 3 exonucleases, including human Exonuclease I (hExoI). HExoI plays several important roles in cells, in processes such as homologous recombination, telomere maintenance, and mismatch repair. We found that cyclo-dA was a complete block to all three enzymes tested (RecJ, lambda exonuclease, and hExoI), whereas 8-oxo-dA was only a partial block. However, under conditions of limited enzyme, including conditions relevant for mismatch repair, both lesions were complete blocks to hExoI. Thus, to the extent that ExoI is necessary for mismatch repair, these results indicate that both cyclo-dA and 8oxo-dA may be endogenous modulators of the mismatch repair pathway. A manuscript describing these results will soon be submitted form publication.[unreadable] [unreadable] Stabilization of the glycosidic bond in cyclo-dA rules out spontaneous glycosidic bond hydrolysis as a mechanism for removing cyclopurines in the absence of NER [unreadable] As described in my review (Brooks, 2007) , the 8,5&#61602;-cyclopurine deoxynucleosides are specifically repaired by the NER pathway, and there are no known back-up pathways for the repair of this lesion. Given the possible role of cyclopurines in XP neurological disease, it was of interest to determine whether there are any other mechanisms for removing such lesion from the genome. One possible mechanism would be spontaneous hydrolysis of the 1&#61602;-9 glycosidic bond followed by enzymic processing. If the presence of the additional 8,5 bond in cyclo-dA resulted in a significant increase in the rate of 1&#61602;-9 glycosidic bond hydrolysis (as has been shown to be the case for other DNA adducts), such a coupled degradation-repair pathway could be of clinical significance. To address this possibility, we developed HPLC and LC-MS methods for the detection and characterization of 5&#61602;(S)-8,5&#61602;-cyclo-2&#61602;-deoxyadenosine containing a hydrolyzed glycosidic bond. Using these methods, we investigated the stability of the glycosidic bond in S-cdA against acid-induced glycosidic bond hydrolysis. However, in contrast to the model described above, we found that the presence of the 8,5 bond in S-cdA actually reduced the rate of glycosidic bond hydrolysis, by a factor of approximately 40-fold. Our observations render any hypothetical mechanism for the removal of 8,5&#61602;-cyclopurine deoxynucleosides from cellular DNA that involves nonenzymic glycosidic bond hydrolysis highly unlikely, and further emphasize the extraordinary chemical stability of these lesions. This work has published (Theruvathu et al 2007). [unreadable] [unreadable] Nuclear Localization of the ATM Protein in Human Purkinje Neurons[unreadable] Another human DNA repair disease is ataxia telangiectasia (AT) and AT-like disorder (ATLD). AT patients develop a progressive cerebellar degeneration in which cerebellar Purkinje neurons are specifically affected. An understanding of the molecular basis of neurodegeneration in AT has been complicated by published work showing that the ATM (ataxia-telangiectasia mutated) protein, which has been shown in numerous other studies to play a crucial role in the detection of DNA damage, is present in the cytoplasm of Purkinje neurons. These observations indicated a novel role for the ATM protein aside from DNA damage sensing in Purkinje neurons. Because of the implications of this result for understanding the mechanistic basis of neurodegeneration in AT patients, we have re-evaluated this question, utilizing postmortem brain tissue from AT patients as a negative control to conclusively establish the specificity of the staining. We found that, it contrast to previous reports, the ATM protein is localized to the nucleus of human Purkinje neurons. Furthermore, although AT and ATLD patients show similar neuropathologies, we found that the location of ATM and MRE11 (which is mutated in ATLD patients) are not identical. Specifically, we found that ATM is present in both the nucleus and nucleolus, whereas MRE11 is excluded from the nucleolus. These observations raise the possibility that ATM plays a role in two distinct pathways: one operative in the nucleoplasm, and an MRE11 independent role in the nucleolus. This work was published in DNA Repair (Gorodetsky et al, 2007).[unreadable] Because Purkinje neurons are specifically affected in human alcoholics, understanding the mechanisms of specific vulnerability of Purkinje neurons in hereditary repair disease may be of relevance to understanding Purkinje neuron degeneration in human alcoholics as well. [unreadable] [unreadable] Finally, we have identified a major DNA damage response pathway that is activated when cells are exposed to acetalehyde, the main carcinogenic metabolite of alcohol in humans.